Case Study
Case Study
Industry.AI boosts wire winding efficiency and quality with its Vision AI product, “Trust.AI”
Industry.AI boosts wire winding efficiency and quality with its Vision AI product, “Trust.AI”
What is Vision AI?
What is Vision AI?
Vision AI integrates digital IP cameras, local or cloud-based servers for computing, machine learning software (ML), and artificial intelligence (AI) to let computers identify and categorize what is occurring in the frames of videos or pictures to provide intelligence that is helpful in retail, manufacturing, construction, security, medicine and other industries.
Vision AI integrates digital IP cameras, local or cloud-based servers for computing, machine learning software (ML), and artificial intelligence (AI) to let computers identify and categorize what is occurring in the frames of videos or pictures to provide intelligence that is helpful in retail, manufacturing, construction, security, medicine and other industries.
Background
Background
One of the world’s most geographically diversified steel producers, based in India and
operating across 26 countries with a presence in over 50 countries. The company’s success
in value-creating initiatives for customers, combined with its fully integrated steel
operations from mining to finished products, enables the company to serve global growth
markets efficiently.
With an annual crude steel production capacity of over 13 MnTPA in India and 12.1
MnTPA in Europe, the company continually strives for operational excellence and product
quality. The company is dedicated to adopting advanced technologies to ensure superior
customer satisfaction and operational efficiency. In this context, the company sought to
resolve production inefficiencies in its high-speed wire winding process using AI-driven
solutions.
One of the world’s most geographically diversified steel producers, based in India and
operating across 26 countries with a presence in over 50 countries. The company’s success
in value-creating initiatives for customers, combined with its fully integrated steel
operations from mining to finished products, enables the company to serve global growth
markets efficiently.
With an annual crude steel production capacity of over 13 MnTPA in India and 12.1
MnTPA in Europe, the company continually strives for operational excellence and product
quality. The company is dedicated to adopting advanced technologies to ensure superior
customer satisfaction and operational efficiency. In this context, the company sought to
resolve production inefficiencies in its high-speed wire winding process using AI-driven
solutions.
Challenges faced during steel manufacturing and
how AI can help:
how AI can help:
Challenges faced during steel manufacturing and
how AI can help:
High-carbon steel wires, ranging in diameter from 0.89 mm to 1.6 mm, are drawn from wire rods and wound onto metallic spools at speeds of 500 m/min. The wire winding process, a critical component of the manufacturer’s production line, faced several challenges
High-carbon steel wires, ranging in diameter from 0.89 mm to 1.6 mm, are drawn from wire rods and wound onto metallic spools at speeds of 500 m/min. The wire winding process, a critical component of the manufacturer’s production line, faced several challenges
Uneven Winding: If the traverse arm stroke setting is misaligned with the spool width, the
winding process becomes uneven, leading to either concave or convex winding shapes.
This inconsistency increases the risk of the wire rings loosening and entangling during
unwinding at customer sites, resulting in product rejection and customer dissatisfaction.
Uneven Winding: If the traverse arm stroke setting is misaligned with the spool width, the
winding process becomes uneven, leading to either concave or convex winding shapes.
This inconsistency increases the risk of the wire rings loosening and entangling during
unwinding at customer sites, resulting in product rejection and customer dissatisfaction.
Spool Wobbling: During the winding process, spool wobbling affects wire tension,
causing loose winding. Operators, who manage multiple winding units, often miss these
issues, especially given the high speeds and complexity of the process.
Spool Wobbling: During the winding process, spool wobbling affects wire tension,
causing loose winding. Operators, who manage multiple winding units, often miss these
issues, especially given the high speeds and complexity of the process.
Manual Adjustments: Operators manually adjust the stroke settings of the traverse arm
4-5 times per cycle to accommodate the gradual expansion of the spool flanges, which
expand by 10-15 mm due to wire load. Despite these adjustments, inconsistencies in
winding can occur, leading to customer complaints and production inefficiencies.
Manual Adjustments: Operators manually adjust the stroke settings of the traverse arm
4-5 times per cycle to accommodate the gradual expansion of the spool flanges, which
expand by 10-15 mm due to wire load. Despite these adjustments, inconsistencies in
winding can occur, leading to customer complaints and production inefficiencies.
These issues caused significant delays, material waste, and production downtime, adversely
affecting operational efficiency and product quality.
These issues caused significant delays, material waste, and production downtime, adversely
affecting operational efficiency and product quality.
To address these challenges, the company partnered with Industry.AI to implement a Camera Vision-based Winding Correction System. This AI-powered solution is designed to enhance precision in the wire winding process by detecting anomalies and automatically
correcting errors in real time.
To address these challenges, the company partnered with Industry.AI to implement a
Camera Vision-based Winding Correction System. This AI-powered solution is designed to
enhance precision in the wire winding process by detecting anomalies and automatically
correcting errors in real time.
Inconsistency Detection: The system monitors the traverse arm’s stroke setting and spool
lengths at both ends to detect inconsistencies.
Inconsistency Detection: The system monitors the traverse arm’s stroke setting and spool
lengths at both ends to detect inconsistencies.
Auto-Correction: It communicates directly with the PLC to auto-adjust the traverse arm
without human intervention, ensuring uniform winding.
Auto-Correction: It communicates directly with the PLC to auto-adjust the traverse arm
without human intervention, ensuring uniform winding.
Wobbling Detection: The system uses computer vision to detect spool wobbling and alerts
the operator immediately to prevent further issues.
Wobbling Detection: The system uses computer vision to detect spool wobbling and alerts
the operator immediately to prevent further issues.
Vertical Arm Monitoring: The system identifies if the vertical arm stops functioning and
triggers an alarm to notify the operator.
Vertical Arm Monitoring: The system identifies if the vertical arm stops functioning and
triggers an alarm to notify the operator.
Industry.AI Approach:
To improve production quality and overcome these challenges, Industry.AI deployed its
Trust.AI product to drive change.
We connected the camera infrastructure across all the factory floor, evaluated the existing
infrastructure capabilities, and put in place a suitable architecture. The Industry.AI team used the existing camera infrastructure, and additionally installed cameras in order to
achieve the use cases required. Post installation of the required infrastructure, a camera
mapping exercise was done to ensure the cameras were in the right angle, position and
the network infrastructure supported the required analytics to be done.
The points below explain the steps taken by Industry.AI while deploying the Trust.AI product:
The points below explain the steps taken by Industry.AI while deploying the Trust.AI product:
-
-
- Data Collection and Ingestion:
• Video feeds from cameras monitor the winding process.
• Data from machine PLCs is periodically collected to support real-time analytics. - Data Preprocessing:
• Video data undergoes normalization (brightness, contrast adjustments), noise reduction, and feature extraction to prepare for analysis. - Computer Vision Techniques:
• Image processing techniques, such as edge detection and pattern recognition, help detect anomalies.
• Machine learning algorithms enhance the system’s accuracy in detecting irregularities in the winding process. - Real-Time Detection and Correction:
• The system continuously monitors the spool and arm movement, detecting uneven winding, spool expansion, and spool wobbling. Once an anomaly is detected, it automatically sends correction signals to the PLC controlling the vertical arm to adjust stroke settings.
- Data Collection and Ingestion:
-
Industry.AI Approach:
To improve production quality and overcome these challenges, Industry.AI deployed its
Trust.AI product to drive change.
We connected the camera infrastructure across all the factory floor, evaluated the existing infrastructure capabilities, and put in place a suitable architecture. The Industry.AI team used the existing camera infrastructure, and additionally installed cameras in order to achieve the use cases required. Post installation of the required infrastructure, a camera mapping exercise was done to ensure the cameras were in the right angle, position and the network infrastructure supported the required analytics to be done.
The points below explain the steps taken by Industry.AI while deploying the Trust.AI product:
-
-
- Data Collection and Ingestion:
• Video feeds from cameras monitor the winding process.
• Data from machine PLCs is periodically collected to support real-time analytics. - Data Preprocessing:
• Video data undergoes normalization (brightness, contrast adjustments), noise reduction, and feature extraction to prepare for analysis. - Computer Vision Techniques:
• Image processing techniques, such as edge detection and pattern recognition, help detect anomalies.
• Machine learning algorithms enhance the system’s accuracy in detecting irregularities in the winding process. - Real-Time Detection and Correction:
• The system continuously monitors the spool and arm movement, detecting uneven winding, spool expansion, and spool wobbling. Once an anomaly is detected, it automatically sends correction signals to the PLC controlling the vertical arm to adjust stroke settings.
- Data Collection and Ingestion:
-
The Key Use Cases included:
-
-
-
-
-
- Detection and Elimination of Uneven Winding:
• The system monitors both ends of the spool, identifying inconsistencies between the stroke setting of the traverse arm and the expanded length of the spool. Upon detecting convex or concave winding, the AI-based algorithm triggers an adjustment, sending signals to the PLC to correct the traverse stroke before the next winding cycle. - Wobbling Detection and Alarm Trigger:
• The camera monitors the flanges of the spool throughout the winding process. The AI tracks the X-axis movement of the flanges. If any variation beyond the threshold is detected, the system raises an alarm to inform the operator. - Vertical Arm Stoppage Detection:
• The system keeps track of the vertical arm’s movement using camera feeds. If the vertical arm stops moving, the system detects this and triggers an immediate alert, allowing operators to intervene.
- Detection and Elimination of Uneven Winding:
-
-
-
-
The Key Use Cases included:
-
-
-
-
-
-
- Detection and Elimination of Uneven Winding:
• The system monitors both ends of the spool, identifying inconsistencies between the stroke setting of the traverse arm and the expanded length of the spool. Upon detecting convex or concave winding, the AI-based algorithm triggers an adjustment, sending signals to the PLC to correct the traverse stroke before the next winding cycle. - Wobbling Detection and Alarm Trigger:
• The camera monitors the flanges of the spool throughout the winding process. The AI tracks the X-axis movement of the flanges. If any variation beyond the threshold is detected, the system raises an alarm to inform the operator. - Vertical Arm Stoppage Detection:
• The system keeps track of the vertical arm’s movement using camera feeds. If the vertical arm stops moving, the system detects this and triggers an immediate alert, allowing operators to intervene.
- Detection and Elimination of Uneven Winding:
-
-
-
-
-
Benefits
Benefits
Increased Efficiency: By automating the detection of inconsistencies and spool wobbling,
the system reduces the need for manual intervention, enhancing production throughput.
This has led to faster cycle times and greater overall efficiency.
Improved Quality Control: Real-time detection of winding anomalies ensures consistent
and accurate winding across all spools, leading to fewer defects and reduced product
rejections at customer sites.
Enhanced Safety: The system’s automated alerts for spool wobbling and vertical arm
malfunctions help operators address issues before they become safety hazards, reducing
the risk of accidents and equipment damage.
Cost Savings: With reduced material wastage and minimized downtime, the system
contributes to significant cost savings. Early detection of issues prevents extended
downtime, while process optimization reduces labor costs.
Data-Driven Decision Making: The centralized dashboard provides real-time monitoring
and historical performance analysis. This data enables The company to fine-tune its winding
processes and improve decision-making based on actionable insights.
Scalability and Flexibility: The system’s modular design allows it to integrate with existing
PLCs and be adapted to other winding units or processes. Its flexibility ensures it can evolve
with the company’s production needs over time.
Increased Efficiency: By automating the detection of inconsistencies and spool wobbling,
the system reduces the need for manual intervention, enhancing production throughput.
This has led to faster cycle times and greater overall efficiency.
Improved Quality Control: Real-time detection of winding anomalies ensures consistent
and accurate winding across all spools, leading to fewer defects and reduced product
rejections at customer sites.
Enhanced Safety: The system’s automated alerts for spool wobbling and vertical arm
malfunctions help operators address issues before they become safety hazards, reducing
the risk of accidents and equipment damage.
Cost Savings: With reduced material wastage and minimized downtime, the system
contributes to significant cost savings. Early detection of issues prevents extended
downtime, while process optimization reduces labor costs.
Data-Driven Decision Making: The centralized dashboard provides real-time monitoring
and historical performance analysis. This data enables The company to fine-tune its winding
processes and improve decision-making based on actionable insights.
Scalability and Flexibility: The system’s modular design allows it to integrate with existing
PLCs and be adapted to other winding units or processes. Its flexibility ensures it can evolve
with the company’s production needs over time.
Results / Conclusion:
The implementation of the Camera Vision-based Winding Correction System in the
company’s wire winding process has yielded significant operational improvements. By
addressing core challenges such as inconsistencies in stroke settings, spool lengths, and
equipment malfunctions, the system has enhanced efficiency, quality, and safety.
Automated real-time monitoring and alerts reduce the need for manual intervention and
empower operators to quickly respond to issues.
Furthermore, the centralized reporting and analytics platform provides data-driven insights, enabling continuous optimization of the winding process. This system not only reduces labor costs, minimizes material waste, and prevents equipment downtime but also improves overall product quality, leading to fewer customer complaints.
A key benefit for the company is the ability to assure its end customers that their winding process is powered by a cutting-edge machine vision system, ensuring a higher standard of product quality and reliability. This positions them as an innovator in the steel industry, enhancing its reputation and offering a competitive edge in the market.
In conclusion, the Camera Vision-based Winding Correction System represents a pivotal advancement in their production capabilities. The solution has demonstrated scalability, flexibility, and adaptability, ensuring its long-term value and contribution to a safer, more efficient, and customer-centric manufacturing process.
Furthermore, the centralized reporting and analytics platform provides data-driven insights, enabling continuous optimization of the winding process. This system not only reduces labor costs, minimizes material waste, and prevents equipment downtime but also improves overall product quality, leading to fewer customer complaints.
A key benefit for the company is the ability to assure its end customers that their winding process is powered by a cutting-edge machine vision system, ensuring a higher standard of product quality and reliability. This positions them as an innovator in the steel industry, enhancing its reputation and offering a competitive edge in the market.
In conclusion, the Camera Vision-based Winding Correction System represents a pivotal advancement in their production capabilities. The solution has demonstrated scalability, flexibility, and adaptability, ensuring its long-term value and contribution to a safer, more efficient, and customer-centric manufacturing process.
Results / Conclusion:
The implementation of the Camera Vision-based Winding Correction System in the
company’s wire winding process has yielded significant operational improvements. By
addressing core challenges such as inconsistencies in stroke settings, spool lengths, and
equipment malfunctions, the system has enhanced efficiency, quality, and safety.
Automated real-time monitoring and alerts reduce the need for manual intervention and
empower operators to quickly respond to issues.
Furthermore, the centralized reporting and analytics platform provides data-driven insights, enabling continuous optimization of the winding process. This system not only reduces labor costs, minimizes material waste, and prevents equipment downtime but also improves overall product quality, leading to fewer customer complaints.
A key benefit for the company is the ability to assure its end customers that their winding process is powered by a cutting-edge machine vision system, ensuring a higher standard of product quality and reliability. This positions them as an innovator in the steel industry, enhancing its reputation and offering a competitive edge in the market.
In conclusion, the Camera Vision-based Winding Correction System represents a pivotal advancement in their production capabilities. The solution has demonstrated scalability, flexibility, and adaptability, ensuring its long-term value and contribution to a safer, more efficient, and customer-centric manufacturing process.
Furthermore, the centralized reporting and analytics platform provides data-driven insights, enabling continuous optimization of the winding process. This system not only reduces labor costs, minimizes material waste, and prevents equipment downtime but also improves overall product quality, leading to fewer customer complaints.
A key benefit for the company is the ability to assure its end customers that their winding process is powered by a cutting-edge machine vision system, ensuring a higher standard of product quality and reliability. This positions them as an innovator in the steel industry, enhancing its reputation and offering a competitive edge in the market.
In conclusion, the Camera Vision-based Winding Correction System represents a pivotal advancement in their production capabilities. The solution has demonstrated scalability, flexibility, and adaptability, ensuring its long-term value and contribution to a safer, more efficient, and customer-centric manufacturing process.